Is Microsoft's 'Singularity' the OS of the future?
Microsoft's TechFest internal science fair wasn't just about social networking and telescopes.
The company also discussed new technology closer to its roots: an operating system kernel concept called "Singularity" intended as a showcase for some cutting-edge computer science.
The software isn't the next version of Windows or a reheated DOS. It's a prototype of an operating system intended for computer science research that Microsoft said demonstrates the possibilities for software that is more dependable and secure than contemporary OSes (yes, that includes Windows).
"Singularity is not the next Windows," Rick Rashid, senior vice president of Microsoft Research, said in a statement. "Think of it like a concept car. It is a prototype operating system designed from the ground up to test-drive a new paradigm for how operating systems and applications interact with one another. We are making it available to the community in the hope that it will enable researchers to try out new ideas quickly."
The Singularity architecture
(Credit: Microsoft Research)If you're itching to take a look at Singularity, Microsoft has made a research development kit available for free download from its CodePlex Web site.
The RDK includes source code, build tools, test suites, design notes, and other background materials. Microsoft says the software, governed by a special Microsoft Research license, can be used for noncommercial, academic projects only.
Microsoft said Singularity has been in development for more than five years. "More than 40 Microsoft Research researchers and interns have collaborated on the project, which incorporated their ideas on security, programming languages, tools, and operating systems--and accelerated their own research," according to a Microsoft Research post describing the project.
The roots of the project stem from research into what a modern operating system would look like and how it would behave. Microsoft points out that Windows, like Unix, Linux and the MacOS, all trace their origins back to Multics, an operating system that originated in the mid-1960s. In essence, the operating systems we use today are built on foundations that are more than 40 years old.
Singularity is written in an extension of C#, Microsoft's high-level programming language, as opposed to C or C++, which typically have been used for current operating systems. By using C#, Microsoft said, the researchers prevented a class of errors known as buffer overruns, "thereby eliminating an area of vulnerability typically exploited by worms and viruses."
Singularity is clearly a research project. It's an operating system kernel without a user interface. Still, as Larry Dignan at ZDNet points out, it does make you wonder whether some folks at Microsoft would like to start over from scratch with the next version of Windows.
Given the driver troubles, legacy issues, and compatibility headaches with Vista, that might not be a bad idea.
Mike Ricciuti joined CNET in 1996. He is now CNET News' Boston-based executive editor and east coast bureau chief, serving as department editor for business technology and software covered by CNET News, Reviews, and Download.com. E-mail Mike.
- Tags:
-
Operating system,
-
Microsoft,
-
research,
-
TechFest
- Share:
- Digg
- Del.icio.us
The download link is here:
http://www.codeplex.com/singularity
CNet disabled hyperlinks so you have to copy and paste it into your web browser URL field.
Get it while Microsoft still allows it to be downloaded, for academic use only. It will eventually replace Windows and make Linux and Mac OSX look like CP/M.
Also, if super restrictive DRM ever got put it in. You'd still have a few looking for an alternative.
You also have to consider someone creating an alternative. Linuxarity maybe? I don't know. We'll never get rid of the OS flame wars. That's for sure.
One shouldn't have to own the latest greatest hardware just to be able to run an OS, it should be able to go back into time and include a range of older hardware and the OS should automatically adjust and still run somewhat smooth and operational.
Linux does this well and OS X too. Anyways, there is a lot of smart people in the world that know how hardware and software works, this is a tease at this point.
I haven't bought Windows since XP, it doesn't deserve my money, if I had any respect for it, I would have purchased it, what once every 5 years ? lol
It works against everything they've done for the last 30+ years and no one is going to see anything useful come from this for a while. Microsoft has said as much in their license and other articles.
Right now the multitude of hobby OSes are more useful than Singularity and they will be for a long time (both in current functionality and availability to the masses).
It's a shame...
At any rate, the concept sounds very cool - a managed kernel and SIPs are two of the things that have really stuck out in my mind as a huge benefit.
/P
Why light a candle to "singularity" when there isn't any real upfront feasibility or real information about it yet.
I mean honestly they were working on this at the same time they were working on Vista? And now this is becoming overhyped.
I was looking forward to Vista but after using it for about a month, let's just say it's quite disappointing. Although Microsoft stands by the product, I can't say I will. For the simple fact they left a lot of things out - the driver support issue was a biggy. It was the kicker for me. It's as if they didn't think it through enough in the 6 six years they had to work on it. And lets just face it, it was six years not five.
And now another OS project is taking five years to be finished but instead of it being longhorn 2 its now called singularity. What a bunch of hooey.
If it wasn't for their entrenched marketshare and obvious monopoly in the OS market. I just don't see how much long they can survive on hype.
This is a RESEARCH PROJECT! This is not Longhorn. This is not the next version of Windows. In fact, this is not even a product! It is a RESEARCH PROJECT. Get it? A RESEARCH PROJECT!
I'll try to explain this in small words so that you can understand it: this is a re-search-pro-ject.
Research is something that you do when you want to explore new concepts. Some of the concepts might eventually make their way into future products at Microsoft or elsewhere. But this, by itself, is not a project.
Multitasking? Gee, at one time it was a concept that was part of a RESEARCH PROJECT. Graphical user interfaces? At one time a RESEARCH PROJECT. Keyboards? Yup, you guessed it, at one time a RESEARCH PROJECT.
Is any of this sinking in yet?
That was Vista, right? Instead of incrementally improve Windows, they redid the UI (aero), the device model, etc. Other things (like the new file system) they had to blow off since even MSFT with its resources couldn't do it all from scratch. But the intent of Vista was to start over, and ironically enough, improved security starting with the kernel was what carried the day. (Ironically since that's the same claim as "singularity"). MSFT thought they could never make XP secure.
Also, if I may remark, I find it ironic that XP is now "the new OS" as MSFT scrambles to find something that will run on sub $300 computers or the like.
Sure, I hope we use electric or hydrogen cars someday, but the internal combustion still functions well after about 100 years. The fact that the kernel of an OS is 30 years old may not doom it. Leopard seems to be doing just fine....
"The fact that the kernel of an OS is 30 years old may not doom it. Leopard seems to be doing just fine...." --- What do you mean? If everyone thought like you did, we'd all still live in caves, because at the time, they were good enough. This isn't just change for the sake of change. Monolithic kernels are... well... UNSAFE. You'll see what I mean soon enough ;)
Conceptually, one wouldn't think this to be true because of the modularity found in the architecture. But... as we see with the bloat-fest called Vista, it turns out to be all too true.
The monolithic kernels have turned out to be far more efficient, far better-built, and far better able to change and adapt (see also OSX, Linux).
I would've been more impressed if MSFT finally embraced an architecture akin to *nix and worked off of that. I thought for awhile that they were doing that with Server 2008, but I guess not... pity, that.
/P
Microkernels weren't efficient before because they needed expensive hardware protection features. Switching address spaces and so on takes a lot of time when using the CPU for it. CIL makes it possible to have a safe environment without using hardware protection features.
'I don't think Vista was a new OS'
was suppose to be a reply to:
'New OS from scratch' post above, it was not suppose to be free standing.
Mmmmhhh... That really sounds like the OS of the future that we now have, and we call LINUX!
Well, duh.
that and the fact this is M$ we are talking about is enough not to trust it. Not unless it is released under the REAL GPL should it even be considered.
A super massive star, turns off, then collapses in on itself creating a black hole or "singularity".
What a concept name - perhaps the name is foreshadowing?
This looks like another vaporware project like the new file system they have been working on for about 12 years now.
Maybe this will be a first for MS. Don't hold your breath.
this project. Hmm. Thats 0.00052% of their workforce.
Looks like they are real dedicated to this project. This is just more
smoke coming Microsoft trying to freeze customers into waiting on
something great, that just maybe will be delivered in 20 years.
you to read the article carefully, and found the obvious short
comings of this "experiment".
I could go on all day if I tried if I just went by the article. I think one should read the source though before we decided if it would pan out based on its technical merits, and I?m just not feeling up to it this morning or tomorrow for that matter.
Most research projects don?t pan out anyway. However they do usually provide smaller ideas that do made it into real products. If it really worked it wouldn?t be a research product. It would just be a product.
So please, I?m interested in exactly what your referring to.
- ISTR...
-
by MLKahnt
March 12, 2008 3:31 PM PDT
- Pascal was originally developed as a programming design expression for academic purposes. Linux was a college exercise, and much of Unix development for its first two decades was an ongoing distributed college computer science exercise. Sometimes it is just best to indulge a period of research and experimentation.
-
Reply to this comment
-
See all 56 Comments >>